超越并行采样：面向智能搜索的多样化查询初始化
Beyond Parallel Sampling: Diverse Query Initialization for Agentic Search
AI总结：
针对智能搜索中的广度缩放，提出DivInit方法，通过在第一轮生成多样化查询而非独立采样，缓解查询冗余问题，在多跳问答任务中平均提升5-7个点。
摘要：
测试时缩放用于智能搜索通常增加深度（即每个轨迹更多轮次和令牌）或广度（即更多并行展开）。这里我们关注广度缩放，表明标准并行采样收益递减，并将其归因于第一轮的查询冗余。当模型在不同展开中发出相似的第一查询时，线程检索重叠的证据，后续轮次基于此共享检索。我们通过DivInit解决这一限制，这是一种在第一轮无需训练的干预。DivInit不是采样k个独立的第一查询，而是从单次调用中抽取n个候选，选择k < n个多样化的种子，并将它们作为并行轨迹运行。在五个开源模型和八个基准测试中，DivInit始终优于标准并行采样，在匹配计算量的多跳问答上平均提升5到7个点。代码可在https://this URL获取。
英文摘要：
Test-time scaling for agentic search typically increases depth (i.e., more turns and tokens per trajectory) or breadth (i.e., more parallel rollouts). Here we focus on breadth scaling, showing that standard parallel sampling yields diminishing returns, tracing this to query redundancy at the first turn. When models issue similar first queries across rollouts, the threads retrieve overlapping evidence, and subsequent turns are conditioned on this shared retrieval. We address this limitation with DivInit, a training-free intervention at the first turn. Rather than sampling k independent first queries, DivInit draws n candidates from a single call, picks k < n diverse seeds, and runs them as parallel trajectories. Across five open-weight models and eight benchmarks, DivInit consistently improves over standard parallel sampling, with average gains of five to seven points on multi-hop QA at matched compute. Code available at https://github.com/cxcscmu/diverse-query-initialization
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
